Recognition using Deep Networks
April 2023
This project’s aim is to build, train, analyze, and modify a deep network for a recognition task (digit/Greek letters). MNIST digit recognition data set is used to build and train the network. The network model is successfully implemented in Python with the help of Pytorch library. It also provides the scope to examine and/or modify my own network using the constructed model. An experiment is also done to evaluate the effect of changing different aspects of the network.
Please find the report for the same below.